The generalized scaling curve method (Patzek et al., 2017) was proposed as a solution to substitute the empirical decline curve analysis that may be inappropriate to fit the production of shale oil wells. As the method is derived from the physics-based model on fractured well geometry, it can match the rapid one over square root of time declines in early production and moderate exponential declines in boundary-dominated production, where the popular decline curve analysis struggles. The proposed method was believed to deliver the results that are comparable to numerical reservoir simulation with less time and data requirement. In this paper, the generalized scaling curve method is validated using field data from Niobrara and Eagle Ford shales. The results are then compared to an analytical solution from an anomalous diffusion approach (Albinali et al., 2016) that has been successfully applied to the data. A synthetic reservoir simulation model is also created and run with the reservoir simulator GigaPOWERS. The simulated oil production rate is then adopted as the synthetic data to feed the generalized scaling curve input. Sensitivity analysis is conducted using the numerical model to investigate the role of porosity, permeability, gas oil ratio, initial reservoir pressure and hydraulic fracture spacing on affecting the scaling curve. The results suggest that the generalized scaling curve is a robust practical approach that is somewhere in between decline curve analysis and numerical reservoir simulation and can be used to predict the EUR of shale oil wells. The method is shown to give similar results to the analytical and simulation solutions with predominantly requiring production data only as the favoured empirical decline curve analysis.